Netherlands Defence Academy

• Dynamic operation of fixed bed reactors is becoming increasingly common • Unsteady 1D models do not capture heat transfer limitations in packed beds • Dynamic instrumental analytical techniques are needed but complex and expensive Recent research in chemical plant operation shows increasing interest in dynamic process operation as part of designed operating strategy for reasons such as increased dependency on renewable energy, and process intensification. Conventional analyses of fixed bed reactors are developed for steady state optimization and may not be adequate for dynamic operation. In fact, the important metrics and targets in dynamic process design are not entirely clear. The first objective of this article is to provide a state-of-the-art survey categorize types of dynamic operation, and rank the available common modelling and analytical tools suitable for quantification of dynamic process variables. The article then examines a case study of 1D and 2D model differences in a methanol steam reforming reactor. The case study shows model prediction differences of up to 15% for conversion, and up to 50% for CO concentration at the outlet during extreme load changes. The study concludes that the complexity of analytical and numerical techniques for dynamic processes is notably higher compared to steady state analyses, but appropriate tools and procedures are currently lacking.

As navies and maritime organisations transition towards low-emission propulsion systems, spark-ignited (SI) gas engines capable of operating on sustainable, low-reactivity fuels are gaining renewed interest. These engines, while offering potential for fossil-free operation, present significant challenges under transient conditions due to complex interactions between throttle control, fuel regulation, and combustion stability. Accurate dynamic modelling is critical to integrate these engines into resilient naval power systems and to support the development of advanced control strategies. This study evaluates several high-fidelity mean value first principle engine modelling (MVFPEM) approaches for simulating the dynamic gas path behaviour of a large, high-speed, SI marine engine under rapid load changes. Models with varying levels of complexity, including simplified and full turbocharger implementations and different gas path volume resolutions, were calibrated using a single measurement campaign and validated against measured transient data. Several methods for turbocharger performance mapping (Stapersma, Casey & Robinson, and Jensen) were evaluated for their applicability in predicting the engine behaviour in dynamic operating scenarios. The results highlight that models incorporating three control volumes and full turbocharger dynamics achieve the highest accuracy, particularly during rapid load increases and recovery phases. Simplified models fail to capture turbocharger inertia and pressure transients, limiting their applicability to investigate naval propulsion or electric power generation plant behaviour under transient load conditions. This work provides guidance on selecting and validating engine models for marine applications and reinforces the role of high-fidelity MVFPEMs in the design and simulation of future naval energy systems.

ABSTRACT Background: Moral injury has garnered increasing attention in recent decades, extending from the military to professions like healthcare and policing. Defined as psychological, social and spiritual suffering resulting from moral transgressions, moral injury is now recognised as a multidimensional phenomenon. A growing body of empirical and theoretical research is establishing a foundational understanding of its psychological, moral and contextual dimensions. Their findings underscore the need to broaden the understanding of moral injuries sustained in the workplace, as well as work-related mental health more generally, by incorporating ethics- and context-informed approaches. This applies to both research and practice. However, the actual practical implications of their findings remain unexplored. Objective: This contribution makes a start in addressing the gap between emerging moral injury research and its practical implementation. It seeks to identify potential interventions and caution against approaches that may inadvertently exacerbate harm. Approach: Drawing on recent interdisciplinary literature, this contribution synthesises key findings and translates them into practical recommendations. It proposes individual, organisational and wider societal levels of intervention, as such with particular attention to ethics-informed and context-sensitive strategies. Results: Research findings point to several promising interventions, including resilience training, ethics education, and moral case deliberation, targeted at both frontline practitioners and leadership. Organisational approaches such as Just Culture initiatives and practices of transformative recognition are also suggested. Conversely, the paper identifies potential pitfalls in popular responses such as the uncritical use of positive psychology or ‘integritism’, which may inadvertently reinforce moral distress. Conclusions: Adequately addressing moral injury in high-impact professions requires a comprehensive, multi-level approach. Cross-disciplinary collaboration is likely to be beneficial, involving mental health professionals, ethicists, chaplains and organisational specialists, to address the various dimensions of moral injury. This contribution has offered several research-informed suggestions for interventions, which nevertheless require further theoretical development and empirical examination.

Cross-view localization between Unmanned Aerial Vehicle (UAV) and satellite imagery is crucial for autonomous navigation in GPS-denied environments. However, large domain gaps, including viewpoint discrepancies, scale variations, and appearance differences — pose significant challenges. In this paper, we propose the Adaptive Multi-scale Cross-modal Fusion Network (AMCF-Net), a novel approach that effectively addresses these limitations through a shared backbone architecture and adaptive fusion mechanisms. Unlike previous dual-backbone approaches that process UAV and satellite images separately, our method employs a unified FocalNet-Tiny backbone to extract cross-modal features, followed by a Spatially-adaptive Crossmodal Feature Fusion (AMCF) module that dynamically combines multi-scale similarities using learned adaptive weights. This shared representation learning enables better cross-modal alignment and significantly reduces computational overhead. Comprehensive experiments on the UL14 benchmark demonstrate that AMCF-Net achieves state-of-the-art performance, with a Relative Distance Score (RDS) of 78.12% and meter-level accuracy of 27.25% at 3 m, 50.16% at 5 m, 84.37% at 10 m, and finally 88.51% at 20 m. Ablation studies further validate the effectiveness of the shared backbone and adaptive fusion mechanism, demonstrating significant improvements over traditional separate processing approaches.

Emerging clean fuels with high octane rating make spark ignition (SI) technology a promising candidate for heavy-duty applications. The conversion of existing diesel engines to SI operation can accelerate the adoption of these fuels. This study investigates the combustion characteristics of a 500 kWe marine lean-burn (LB) homogeneous charge SI engine with a flat cylinder head and a hemispherical bowl-in piston. It focuses on the relationship between fuel distribution and phasing across the distinct bowl-in and squish combustion phases and their impact on efficiency and emissions in multicylinder engines. The effects of air excess ratio, spark timing, and intake air temperature are systematically assessed. Dedicated measurements of methane and total unburned hydrocarbon emissions enable a comprehensive evaluation of combustion performance and emissions. Results confirm the presence of a slower squish phase, differing from conventional SI engines, and highlight the influence of the squish region’s surface-to-volume ratio on flame propagation. The sensitivity of combustion behavior to control parameters such as air excess ratio and ignition timing is demonstrated, with notable differences: while richer mixtures advance bowl-in and squish phases, earlier ignition timing delays the squish phase. Despite this, both mixture enrichment and ignition timing advancement improved performance, increasing brake thermal efficiency by 25% and 10%, respectively. Methane emissions remained within typical ranges for marine SI engines and NO x emissions met Tier III limits at nominal conditions; yet the persistent challenge of methane slip underscores the need for more comprehensive regulatory standards addressing both CH 4 and NO x emissions. • Dual-phase (bowl-in and squish) combustion characterized in a 500 kWe marine lean-burn SI engine. • Quantifies impact of combustion phasing and spark timing on efficiency, stability, and emissions. • Methane and NO x emissions evaluated against IMO Tier III and discussed for regulatory context. • Findings support optimization and retrofitting strategies for sustainable marine engine applications.

Europe's Critical Underwater Infrastructure (CUI) is both economically vital and vulnerable to sabotage. Recent incidents in the Baltic Sea suggest a deliberate strategy to exploit these vulnerabilities, by dragging anchors across seabed cables. The vast spatial extent of CUI makes such attacks difficult to detect and deter. Distributed Acoustic Sensing (DAS) enables continuous monitoring by detecting vibrations along submarine fiber-optic cables, offering maritime situational awareness. These vibrations are caused by surface vessels, submarines, or unmanned underwater vehicles. To enable real-time monitoring, we combined publicly available DAS and AIS datasets to label ship signals in the DAS data. Using this dataset, we trained a convolutional neural network (CNN) to detect vessels in large volumes of DAS data. We evaluated the CNN using two datasets, two data processing techniques, and augmentation strategies. Our CNN effectively distinguishes weak vessel signals from substantial background noise, demonstrating its potential for real-time maritime threat detection. Based on our findings, we offer recommendations for applying DAS-based monitoring to protect CUI in the North Sea.

AI-enabled object detection and classification is being developed and deployed by militaries worldwide. In the maritime domain, autonomous vehicles such as unmanned aerial vehicles (UAVs) and unmanned surface vehicles (USVs) offer new possibilities for obtaining maritime situational awareness. However, these technologies also introduce new threats for our vessels: they can be deployed to detect and classify naval vessels. To counter this threat, we developed anti-AI camouflage for naval vessels. Using open-source publications from the Dutch Ministry of Defence, we compiled a dataset of Dutch Royal Navy vessels, consisting of five classes. The images, captured from the viewpoint of USVs and UAVs, were used to design camouflage patterns - adversarial patches - intended to fool AI-enabled object detection. These patterns reduce the likelihood of AI detecting naval vessels or cause the AI to misclassify the type of vessel. By picking the colors of the pattern, we achieve dual-attribute camouflage that can fool both the human and artificial brain.

Abstract Rapid integration of Artificial Intelligence (AI) into the military domain necessitates actionable strategies for translating high-level principles of responsible use into practical guidelines. However, there remains a problematic gap between these principles and the norms that govern the use of AI in military operations. Moreover, these norms are highly dependent on the particular context in which military AI is deployed. This leads to normative uncertainty; what is responsible use of AI in a specific military operation? Unclear practical guidelines pose challenges for technology developers and military operators involved in the deployment of military AI. This paper emphasises the need for a context-specific assessment of responsible use of military AI. Moving beyond a one-size-fits-all standard, we propose the Military AI Responsibility Contextualisation (MARC) framework; a structured approach that facilitates a context-specific assessment. In that way, this paper aims to contribute to bridging the gap between abstract principles and practical guidelines. We furthermore emphasise the need for interdisciplinary collaboration in further operationalizing responsible military AI to work towards the ethical and effective development and deployment of AI in military operations.

Transport pipelines are critical for the distribution of energy commodities, refined products and essential materials like water and hydrogen.While often buried underground for protection from environmental factors, they remain vulnerable to distant threats such as intentional or accidental explosions.This study investigates the mechanical response of steel pipelines to seismic excitation from nearby buried explosive charges.An experimental approach was employed to measure the stresses and deformation of a buried, straight steel pipeline segment using high-frequency strain gauges attached to its surface before burial in damp, sandy soil.The gauges were strategically placed to capture various deformation modes, such as ovalisation and axial strain.Seismic loading was applied using buried plastic explosive charges weighing up to 16 kg TNT-equivalent, detonated at distances up to 8 m from the pipeline centre.Seismic ground shock characteristics were evaluated using triaxial MEMS accelerometers.The results establish a relationship between the explosion intensity, measured ground shock characteristics (e.g., peak particle velocity) and pipeline deformation.For the experimental scenarios, pipeline deformation remained within the elastic limit, with maximum strains measured in the order of 100 -200 10 -6 .The ground shock was found to be significantly damped between the detonation point and the pipeline.Thus, when no other loads are present, transport pipelines buried in dry, sandy soil are predicted to deform elastically for subsoil detonations up to 100 kg TNT-equivalent at a 14.5-m distance.